# Analyze Claude Flow Swarm Performance

Performance work is hard when swarm metrics, profiles, and benchmarks are scattered. This skill gives Claude, Codex, and Claude Code a structured analysis workflow.

## Install

```bash
npx skillstore add dnyoussef/when-analyzing-performance-use-performance-analysis
```

## Metadata

- Status: approved
- Slug: dnyoussef-when-analyzing-performance-use-performance-analysis
- Version: 1.0.0
- Author version: 1.0.0
- Skillstore revision: r1
- Version status: valid
- Tree hash: 9e5d78499cb12bd47447747925bd3dfcaf84f488dd3bf5b8b7d66c5121f86b90
- Author: DNYoussef
- GitHub username: DNYoussef
- License: MIT
- Repository: https://github.com/DNYoussef/ai-chrome-extension/tree/main/.claude/skills/performance/when-analyzing-performance-use-performance-analysis
- Ref: c579efd02184f1128d6741d16e6c8d951d7d4208
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/dnyoussef-when-analyzing-performance-use-performance-analysis
- Manifest: https://skillstore.pages.dev/api/skills/dnyoussef-when-analyzing-performance-use-performance-analysis/manifest

## Capabilities

- Establishes baseline metrics for Claude Flow swarms.
- Profiles swarm execution, agents, memory, CPU, and network behavior.
- Classifies bottlenecks by severity and likely impact.
- Guides optimization planning for topology, allocation, caching, and memory use.
- Validates improvements with comparison benchmarks and target checks.

## Use Cases

- Diagnose Slow Swarm Workflows: Find likely bottlenecks when a Claude Flow swarm has poor throughput or high latency.
- Validate Optimization Results: Compare baseline and optimized runs to show whether performance targets were reached.
- Prepare Repeatable Benchmarks: Create a consistent performance workflow for profiling, reporting, and regression checks.

## Prompt Templates

### Create a Baseline

```
Use this skill to establish a baseline for my Claude Flow swarm. Summarize throughput, latency, resource use, and next steps.
```

### Profile a Slow Workflow

```
Use this skill to profile a Claude Flow workflow with slow task completion. Identify likely CPU, memory, network, and coordination bottlenecks.
```

### Compare Optimization Results

```
Use this skill to compare baseline and optimized benchmark results. Rank bottlenecks and recommend changes with validation metrics.
```

### Run a Full Improvement Cycle

```
Use this skill for a full performance improvement cycle on a production-like swarm. Require approval before any state-changing optimization command.
```

## Limitations

- Requires Claude Flow and related shell tools to be installed.
- Uses command examples that need review before execution.
- Assumes expected JSON output from the Claude Flow CLI.
- Does not independently verify results from external profiling tools.

## Best Practices

- Establish a baseline before changing topology, memory settings, or agent allocation.
- Review every generated shell command before running it in a shared environment.
- Validate changes with repeated benchmarks and compare results against explicit targets.

## Anti Patterns

- Do not auto-apply optimizations before reviewing recommendations and rollback steps.
- Do not compare results from different workload sizes or measurement windows.
- Do not treat profiler output as complete without checking logs and error rates.

## Security Audit

- Audited at: 2026-07-05T13:39:18.091\+00:00
- Summary: The static backtick and command-substitution findings are mostly Markdown examples or safe shell expressions. A real risk remains because the skill directs runtime npx execution from the alpha channel and includes state-changing optimization commands.

## Stats

- Views: 241
- Downloads: 8
- Favorites: 0
- Popularity score: 0